Development of a new hand-held type thyroid monitor using multiple GAGG detectors for young children following a nuclear accident

Radiation Measurements 150 (2022) 106683

Contents lists available at ScienceDirect

Radiation Measurements
journal homepage: www.elsevier.com/locate/radmeas

Development of a new hand-held type thyroid monitor using multiple
GAGG detectors for young children following a nuclear accident
Kazuaki Yajima a, Eunjoo Kim a, *, Kotaro Tani a, Masumi Ogawa a, b, Yu Igarashi a,
Munehiko Kowatari a, Osamu Kurihara a
a

National Institutes of Quantum and Radiological Science and Technology, National Institute of Radiological Sciences (QST− NIRS), 4-9-1 Anagawa, Inage-ku, Chiba,
263-8555, Japan
Self-Defense Forces, Central Hospital, 1-2-24 Ikejiri, Setagaya-ku, Tokyo, 154-0001, Japan

b

A R T I C L E I N F O

A B S T R A C T

Keywords:
Thyroid exposure
Radioiodine
Thyroid monitor
GAGG detector
Nuclear accident

One of the most important lessons learned from the Fukushima Daiichi Nuclear Power Plant accident is that

direct (in vivo) measurements of thyroid exposure to radioiodine (mainly, 131I) in the affected populations should
be initiated in a timely manner. Furthermore, the existing commercial detectors are not necessarily suitable for
the measurement of young children, who are especially vulnerable to radiation exposure and therefore important
to screen. This paper presents the development of a new thyroid monitor for such measurements and the results
from phantom-based experiments. The monitor has two unique probes having multiple high-energy-resolution
type Gd3(Al,Ga)5O12(Ce) (GAGG) detectors that can be placed directly on the young subject’s anterior neck.
The crystal size of the GAGG detector is 1 cm3, and a probe consisting of a 4 × 1 or 4 × 2 detector array can be
selected depending on the subject’s body size. The thickness of the 4 × 1 array probe is 24 mm, which is less than
that for a conventional NaI(Tl) survey meter having a crystal 1 inch in diameter and 1 inch long (38 mm).
Experimental and computational calibrations of the new monitor using existing and virtual phantoms allowed us
to determine the full energy peak efficiency for the 131I thyroid contents of different age groups from 3-mo-old to
10-yr-old and minimum detectable activity (MDA) values under various conditions. As a result, the attainable
MDA for subjects age ≤5 years under a normal background level (~0.05 μSv h− 1) was found to be ~30 Bq, which
was low enough to identify children with thyroid-equivalent doses over 10 mSv up to about 25 days after the 131I
intake. Our new monitor would be useful in direct thyroid measurements for vulnerable young children
following a large nuclear accident.

1. Introduction
In the case of a nuclear accident with a large release of the radio­
nuclides into the environment, the main concern is exposure to the
residents living in the vicinity of the accident site. In particular, atten­
tion should be paid to the internal thyroid exposure to children due to
intake of radioiodine, as demonstrated in the Chernobyl nuclear acci­
dent in 1986 (Cardis et al., 2011). The largest contributor to the internal
thyroid dose is 131I, with a physical half-life of about 8 days. The dose
delivered following intake of radioiodine is localized in the thyroid, a
relatively small organ in the human body. The thyroid dose per unit of
131
I intake (Sv Bq− 1) is much higher in young children than in adults, in
accord with the thyroid mass defined in each age group’s anatomical

model (e.g., 1.78 g for 1-yr-olds, 20.0 g for adults) (ICRP 1995,

International Commission on Radiological Protection, 1998). The actual
thyroid doses would be less age-dependent in the same exposure con­
dition because inhalation or ingestion intake amounts are smaller for
younger people; however, it should be noted that the radiosensitivity of
the thyroid per unit dose (e.g., the excess relative risk: ERR) was shown
to be high in young children by epidemiological studies related to the
Chernobyl accident (Brenner at al., 2011).
In the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident that
occurred in March 2011, the number of direct thyroid measurements to
determine individuals’ 131I thyroid contents was very limited in contrast
to that in the Chernobyl accident (Kim and Kurihara 2020). Therefore,
the assessment of thyroid doses to most of Fukushima prefecture’s res­
idents had to be performed using other data, such as whole-body counter
measurements targeting 134Cs and 137Cs, radiation monitoring results of

* Corresponding author.
E-mail address: (E. Kim).
/>Received 12 July 2021; Received in revised form 1 November 2021; Accepted 17 November 2021
Available online 19 November 2021
1350-4487/© 2021 The Authors.
Published by Elsevier Ltd.
This is an open
( />
access

article

under

the

CC

BY-NC-ND

license


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

food and drink items, atmospheric transport and dispersion model
simulations for the released radionuclides, and so on (Hosoda et al.,
2013; Kim et al., 2016; Ohba et al., 2020). Great effort has been
expended on these dose assessments, which found that residents’ thy­
roid doses were very unlikely to exceed a thyroid-equivalent dose (TED)
of 100 mSv and were mostly less than 20–30 mSv, even for children in
the relatively high-dose regions (Kim and Kurihara 2020). However, the
actual thyroid doses to individuals can be confirmed only from direct (i.
e., in vivo) measurements, despite the existence of uncertainties about
the nature of radiation measurements, the intake scenario, and so on.
Ecological-model approaches using environmental data (e.g., air con­
centration, ground deposition density) involve still larger uncertainties
and need to be verified by results of direct measurements. This is one of
the critical lessons learned from the FDNPP accident: namely, individual
photon detection to assess thyroid exposure to the possibly affected
populations must be performed in a timely manner so as not to miss the

window for evidence of 131I intake in the case of a major nuclear acci­
dent event.
One technical problem in direct thyroid measurements is that
existing photon detectors are not necessarily appropriate for measure­
ments of young children or newborns (Broggio et al., 2019). We found
that a NaI(Tl) survey meter with a probe, in which a crystal of 1 inch
diameter and 1 inch length is installed, was too bulky to be placed in
close proximity to the anterior neck of preschool children (age <4
years), possibly resulting in decreased sensitivity and reproducibility of
the measurements. This device (TCS-171/172, Hitachi, Ltd., Japan) has
been widely used in Japan for environmental monitoring and was also
intended to be used for the screening of thyroid exposure to radioiodine
in case of nuclear emergencies (JAERI 1993). Indeed, it was used in the
FDNPP accident (Kim et al., 2012). To solve this problem, we proposed
the concept of a new thyroid monitor suited for young children making
use of multiple small photon detectors (Yajima et al., 2020). We selected
high-energy-resolution type Gd3(Al,Ga)5O12(Ce) (hereinafter, GAGG)
detectors, each equipped with a 1-cm cube-shaped crystal; the resulting
monitor performed spectrometric measurements in contrast to the
count/dose rate measurements made with the NaI(Tl) survey meter. Our
previous study used age-specific thyroid phantoms developed by the
Institut de Radioprotection et de Sûret´
e Nucl´eaire (IRSN) (Beaumont
et al., 2017) to compare the minimum detectable activity (MDA) values
for 131I of two arrangements of 8 GAGG detectors with those of NaI(Tl)
spectrometers with larger crystal sizes, and found equal or better per­
formance by the former. This paper introduces our newly developed
thyroid monitor for young children and its performance.

aluminum casings to reduce the probe size. As a result, the exterior

dimension of the casing was 16 mm wide, 17 mm deep, and 35 mm long.
A multi-signal process unit (Signal processor 80406/81416, Clear-Pulse
Co., Tokyo Japan) was used to deal with pulse height spectra from the
GAGG detectors. The electricity was supplied via a Universal Serial Bus
(USB) cable from a laptop equipped with originally developed software
packages for system control, spectrum analysis and data registration.
The main advantages of the GAGG detectors are twofold. First, the size is
small enough for direct thyroid measurements of young children. This is
achieved by the use of a silicon photomultiplier instead of the conven­
tional photomultiplier tube typically used in NaI(Tl) or other scintilla­
tion detectors. Second, the auto-adjustment function of gain-shift due to
the variation of temperature allows us to minimize the need for frequent
energy calibration.
Based on the results of the previous study and mock-up experiments
using manikins or child volunteers and temporal probes (Fig. 1), we
decided upon two detector arrangements for the anterior neck probe: a 4
(row) × 1 (column) array and a 4 × 2 array. We also interviewed two
pediatricians about how to safely examine young children using the
probe. As a result, we obtained the following valuable comments.
• Two subject postures, sitting and supine, are possible. For the use of
the seated position, the subject sits on a small chair or on his/her
mother’s legs while being hugged (although mothers may not always
be the person in charge). The measurer sits face-to-face to the subject
and places the probe on their anterior neck. When the supine position
is used, the subject lies on a bed or is held by his/her mother, who
can place the probe on the subject’s neck herself.
• It will be challenging for young children, in particular those aged 2–5
years old who are likely to move actively, to remain still for the
duration of the measurement time, a few minutes at least. In the
sitting posture, the mother would be able to hold the subject’s head

gently if necessary. Using a toy or something to attract the child’s
attention may also allow the subject to be measured with greater
ease. In any case, the mother’s presence and participation are the
best way for the child to be reassured in such an unusual situation.
Considering the above comments, we developed the new hand-held
thyroid monitor as demonstrated in Fig. 2. This thyroid monitor can
select one of the two detector arrangements described above, and its
weight is about 1 kg (without the laptop), which is lighter than the NaI
(Tl) survey meter (1.5 kg for TCS-171/172). The thickness of the probe is
24 mm in the case of the 4 × 1 detector array with a plastic cover, which
is less than the thickness of the NaI(Tl) survey meter’s probe (38 mm).
The structures of the two probes of the developed thyroid monitor are
shown in Fig. 3. In this illustration, the detectors are arrayed along the
outer surface of the IRSN 10-yr-old phantom with a diameter of 88 mm.
The two probes are currently meant to be used separately depending on
the subject’s age; the 4 × 1 probe for subjects aged ≤5 years old and the
4 × 2 probe for subjects aged >5 years old.

2. Development of the new thyroid monitor
The system for the new thyroid monitor has been described in our
previous paper (Yajima et al., 2020). It uses commercially available
GAGG detectors (Model: HR-GAGGG1C1C1C-type, Clear-Pulse Co.,
Tokyo Japan) enclosed in originally designed and manufactured

Fig. 1. Scenes of mock-up experiments performed to identify the optimal detector arrangements for the new thyroid monitor.
2


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

Fig. 2. Exterior views of the new thyroid monitor. (Panel A: the probe for 4 × 1 detector array, Panel B: the probe for 4 × 2 detector array). The dimensions are the
same in the two pictures.

radioactivity.
We also performed a computational calibration using age-specific
mathematical phantoms (3-mo-old, 1-yr-old, 5-yr-old, and 10-yr-old)
to cover broader age groups. The mathematical phantoms have been
described elsewhere (Ulanovsky and Eckerman, 1998) and were kindly
supplied by Dr. A.V. Ulanovsky at our request. It is noted that the
thyroid-shaped containers in the IRSN phantoms were designed with
reference to the configurations of these mathematical phantoms. The
computational calibration was performed only for the detector
arrangement of the 4 × 1 array through a series of simulations using the
Monte Carlo N-Particle® code ver. 6 (MCNP6) (Werner, 2017). In the
simulations, the internal structure of the GAGG detectors was modeled
for the thyroid monitor, and the number of photons emitted was set to be
large enough to reduce the relative standard deviation to <1%. We also
evaluated the FEP values in a manner similar to that used for the
experimental calibration, by using the pulse height tally of the MCNP6.
A reference location of the probe against each age-specific phantom was
determined in a manner similar to that used in the experiments
(described later) except for the 3-mo-old phantom, for which it was
necessary to model the head tilting backward for probe placement. Fig. 4
shows the simulation models for the probe (4 × 1 detector array)
together with the 3-mo-old, 1-yr-old, and 5-yr-old mathematical
phantoms.
3.2. Efficiency variation with the probe’s displacement
Fig. 3. Structures of the two probes of the new thyroid monitor (left: the probe

for 4 × 1 detector array, right: the probe for 4 × 2 detector array). The lower
figures are probes with thicknesses of 24 mm and 44 mm, respectively.

The probe of the new thyroid monitor is designed to be placed on the
subject’s anterior neck during measurements. However, some gap could
be generated between the probe and the neck in children with excessive
movement. Some deviation of the probe from the location at the
beginning of the measurement is also possible, although the exact probe
location cannot be identified in practice because of the wide interindi­
vidual difference in the thyroid volume (or shape) and a possible
inhomogeneous radioactivity distribution in the thyroid. Thus, we
experimentally obtained FEP values for different locations of the probe
against the IRSN phantoms. A reference location of the probe on each of
these phantoms was decided upon so that the center of the probe was
aligned with the top of the isthmus of the thyroid-shaped container in
the phantom. The displacement of the probe from the reference location
for each phantom was arranged as follows: 0–3.0 cm for the distance
between the probe and the phantom, − 1.0 to 1.0 cm in the vertical di­
rection, and 0◦ –20◦ in the rotational angle (the last two displacements
were the setting on the phantom). We assume that measurements with
the monitor would be reattempted when the displacement of the probe
exceeds these ranges.

3. Calibration and minimum detectable activity
3.1. Counting efficiency
The new thyroid monitor was calibrated using the age-specific thy­
roid phantoms developed by IRSN (Beaumont et al., 2017). Two phan­
toms imitating a 5-yr-old and a 10-yr-old were used in this study. Pulse
height spectra were obtained by the thyroid monitor with a setting of
about 1 keV per channel (1024 channels in total). In the calibration

using the IRSN phantoms, we installed portions of 131I standard solution
(Code: IO010, Japan Radioisotope Association, Tokyo Japan) into the
thyroid-shaped container in each phantom and determined the net peak
area at the primary peak line (365 keV, 81.7%) using a software package
for spectrum analysis (Prime. Advanced Fusion Technology. Co. Ltd.,
Tokyo Japan). The radioactivity in the phantom was adjusted to avoid
problems related to the counting rate (e.g., the dead time above a few
percentages and the energy shift due to the pile-up). We then calculated
the full energy peak (FEP) efficiency values for the 131I thyroid contents
by dividing the net peak areas by the counting time and the loaded

3.3. MDA
The minimum detectable activity (MDA) values of the thyroid
3


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

4. Results
4.1. Counting efficiency
Fig. 5 compares the FEP efficiency values for the 131I thyroid content
in the different phantoms described earlier. Note that the probe was
placed at the reference location on the phantom in all cases reported.
The error bars of the experimental results (for the IRSN phantoms)
denote the combined uncertainty (k = 1) due to the variation of threetimes repeated measurements and the uncertainty of the radioactivity
of the 131I standard solution (1.6%, k = 2). As described earlier, the
experiments were performed for two IRSN phantoms (5-yr-old and 10yr-old), whereas the simulations were performed for the four phan­
toms aged ≤10-yr-old.

The simulation results agreed within 10% with the experimental
results using the 4 × 1 detector array for both the 5-yr-old phantom and
the 10-yr-old phantom, suggesting that our simulations were likely
reasonable for the other phantoms as well. The detector model in the
simulations was validated through benchmark experiments using a133Ba
point source placed at different locations close to the center of the probe
for the 4 × 1 detector array; the discrepancy in the FEP efficiency at 356
keV (62.1%) between the experiments and the summations was
observed to be within 5%. One notable finding in the simulation results
was that the FEP efficiency value for the 4 × 1 detector array was less
influenced by the differences in the phantoms for ages ≤5 yr. This im­
plies that the correction of the FEP efficiency value against the body size
or the subject’s age can be simplified when measuring young children.
On the other hand, the experiments demonstrated that the FEP effi­
ciency of the 4 × 2 detector array was almost double that of the 4 × 1
detector array in the case of the 5-yr-old phantom and the 10-yr-old
phantom.
4.2. Efficiency variation with the probe’s displacement
Figs. 6–8 present the experimental results for the IRSN phantoms (5yr-old and 10-yr-old) to confirm the variation of the FEP efficiency with
the positional displacement of the probe from its reference location on
the phantom. The error bars in these figures have the same meaning as
described above. Fig. 6 displays the relative FEP efficiency (to that in the
case of the zero distance) as a function of the distance between the probe
and the phantom. The FEP efficiency was decreased with increasing
distance, and its relative values were 0.75–0.80 at the distance of 5 mm.
Fig. 7 displays the variation of the relative FEP efficiency with the ver­
tical displacement of the probe. Here, a positive (negative) displacement
denotes that the probe moves upward (downward) from the reference
location. As shown, the FEP efficiency was found to be an asymmetric

Fig. 4. Simulation models of the probe for a 4 × 1 detector array together with
mathematical phantoms for 3-mo-old (the upper), 1-yr-old (the middle) and 5yr-old (the bottom) in the same scale. The right figures are the horizontal views
of each phantom at the level of the dotted lines in the left figures.

monitor for the 131I thyroid content were calculated by the following
equation (Currie 1968; Gilmore 2008):
√̅̅̅
2.71 + 4.65 B
MDA =
ε⋅t
where ϵ is the FEP efficiency at 365 keV of 131I, B is the background
count for the corresponding energy interval, and t is the measurement
time (which was the same between the sample and the background in
this study). We experimentally determined MDA values of the thyroid
monitor with different locations of the probe and under elevated back­
ground radiation levels (up to 2.5 μSv h− 1) by using several 137Csshielded sources placed around the thyroid monitor. This upper back­
ground radiation level was due to the limitation of our experimental
setup. However, this level would be within a reasonable range for the
practical use of the monitor, based on the experiences from the FDNPP
accident where the ambient dose rate was elevated to a few μSv h− 1 or
more at various places in Fukushima Prefecture (including the munici­
palities where prompt evacuation orders were not issued) (Kim et al.,
2020).

Fig. 5. Comparison of the FEP efficiency for the
different phantoms.
4

131

I thyroid contents in


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

Fig. 6. Relative FEP efficiency of the two probes for the 131I thyroid contents in
two IRSN phantoms (5-yr-old and 10-yr-old) as a function of the distance be­
tween the phantom and the probe.

Fig. 7. Variation of the relative FEP efficiency due to the vertical displacement
from the reference location. (Panel A for 4 × 1 detector array and Panel B for 4
× 2 detector array).

profile in the vertical direction. This variation was smaller in the 4 × 2
detector array than in the 4 × 1 detector array. Fig. 8 displays the
variation of the relative FEP efficiency with the rotation angle of the
phantom. The results are provided only for one direction because the
IRSN phantoms have bilateral symmetry. As shown, the relative FEP
efficiency decreased to 0.9 when the phantom was rotated by 20◦ .

seen in recent publications (IAEA 2013; Youngman 2013). A series of
papers on the Child and Adult Thyroid Monitoring After Reactor Acci­
dent (CAThyMARA) project (Broggio et al., 2019) described critical
reviews and recommendations for direct thyroid measurement of the
public. They also address the necessity of having suitable spectrometric
devices for young children (less than 5-yr-old or around) with a contact
geometry to the neck. Our new thyroid monitor appears to meet the
requirement on this technical issue. The thickness of the probe for the 4

× 1 detector array (24 mm) was less than the length of the neck modeled
in the mathematical phantom for a newborn with a tilted head (31 mm)
(Khrutchinsky et al., 2012). As a result, this probe could be placed on the
anterior neck in the simulation model (Fig. 3). Fig. 11 displays scenes of
mockup measurements by the new monitor with the use of a manikin
imitating a newborn infant, suggesting that the contact geometry would
be achievable for real subjects. However, this needs to be confirmed in a
future study through experiments on human volunteers. The manikin
shown in the figure was a commercial product for a nursing practice; its
height is about 50 cm.
In the selection of the detector for the new thyroid monitor, we
prioritized the practicability as a system rather than the performance. In
terms of the energy resolution, cadmium zinc telluride (CdZnTe) de­
tectors have shown better performance compared to scintillation de­
tectors including those developed in recent years; e.g., cerium bromide
(CeBr3), strontium iodide (SrI2(Eu)), and lanthanum bromide (LaB­
r3(Ce)) scintillation detectors (Nishino et al., 2020). However, CdZnTe
detectors are the most expensive when one compares these detectors
with the same sensitive volume, and such a high cost would be a major
obstacle for the mass production of a new thyroid monitor. CeBr3 and
SrI2(Eu) scintillation detectors have better energy resolution compared
to conventional NaI(Tl) detectors (Nishino et al., 2020; Hosoda et al.,
2019). The energy resolution of our newly developed thyroid monitor is

4.3. MDA
Fig. 9 presents the MDA values of the 4 × 1 and 4 × 2 detector arrays
for the 131I thyroid content in two IRSN phantoms (5-yr-old and 10-yrold) as a function of the measurement time under a normal back­
ground condition (~0.05 μSv h− 1). The MDA values in the case of a
counting time of 180 s were around 30 Bq. Fig. 10 presents the MDA
values of the same arrays for two IRSN phantoms (5-yr-old and 10-yrold) as a function of the surrounding ambient dose rate, H*(10), in the

case of a measurement time of 180 s. As a result, the MDA value (1
standard deviation of three-times measurements) increased from 31(6)
Bq at the normal background level of ~0.05 μSv h− 1 to 229(3) Bq at 2.5
μSv h− 1. These MDA values can be applied to the younger age groups (3mo-old and 1-yr-old) because the FEP efficiency values of these groups
were comparable or slightly higher than that of the IRSN 5-yr-old
phantom (Fig. 5).
5. Discussion
It is of great importance to perform direct thyroid measurements of
individuals without delay in the case of a major nuclear accident. This
task would be a great challenge for most of the countries where com­
mercial nuclear power plants are in operation. The significance of the
measurements seems to have been reinforced by the FDNPP accident as
5


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

Fig. 8. Variation of the relative FEP efficiency due to the lateral displacement
of the reference location (Panel A for 4 × 1 detector array, Panel B for 4 × 2
detector array).

Fig. 9. MDA for the 131I thyroid contents as a function of the counting time
(Panel A for 4 × 1 detector array, Panel B for 4 × 2 detector array).

almost the same as that demonstrated in our previous study (Yajima
et al., 2020) (data not shown here) and is comparable to that of the NaI
(Tl) spectrometers, i.e., 7%–8% for the full width at half maximum
(FWHM) at 662 kev. According to the manufacturer’s information, the

energy resolution of the GAGG detector used is 5.1% at 662 keV. We thus
speculate that the energy resolution was degraded because of the use of
the original detector casing and multiple detector arrays. However, this
would not be a significant problem when measuring 131I. The main ra­
dionuclides that were identified soon after the FDNPP accident were
131 132
I, Te–132I, 134Cs and 137Cs (Kurihara et al., 2012), and the adjacent
peak lines (with emission yields >10%, photon energy >100 keV) to the
primary peak line of 131I (365 keV, 81.7%) were from 132Te (228 keV,
88.0%) and 132I (523 keV, 16.0%). Our new thyroid monitor thus has
sufficient energy resolution to observe each of these three peak lines on a
pulse height spectrum without interference from the other two peak
lines. The advantages of the GAGG detectors are described above in
Section 2. In particular, the significant gain-shift due to the variation of
temperature at a measurement place can be a common problem in
scintillation detectors. The GAGG detectors used in the new monitor
solve this problem by fine gain adjustments based on characteristic tests,
and the detectors are also reliable for the use of new thyroid monitor.
Our new monitor is intended to be used with a contact geometry
against the neck; however, it is important to examine the measurement
uncertainty due to the probe’s likely displacement, especially as the
target subjects are young children. This study examined one of the pri­
mary uncertainty factors, the variation of the FEP efficiency due to the
probe’s displacement. As demonstrated in Fig. 6− 8, this variation was
found to be relatively small for the vertical/lateral displacements, but
was expected to be rather significant if there is a continuous gap

between the probe and the neck during measurements of subjects. We
estimated that the uncertainty due to the probe’s displacement is within
~20% in practical use (i.e., assuming a gap of a few millimeters between

the probe and the subject’s neck and some other minor displacements).
Larger displacements of the probe would appear to be abnormal by eye
and were not assumed here. However, actual measurements will also be
accompanied with the counting statistics, the measurement uncertainty
due to the inter-individual differences in the thyroid volume and
thyroid-overlying tissue thickness (Likhtarev et al., 1995; Ulanovsky
et al., 1997), and so on. To reduce the measurement uncertainty, the
counting distance needs to be maximized, as suggested by the previous
studies (Kramer and Crowley, 2000; Beaumont et al., 2018); however,
there is a trade-off relationship with the sensitivity. A long counting
distance is beneficial if the subject’s body has been heavily exposed to
radionuclides (IAEA, 1988), in particular in the case of direct mea­
surements at the thyroid, in which the exact location of this small organ
is not visible from outside the body. The experiences gained from the
FDNPP accident suggest that situations in which members of the public
are overexposed in a nuclear accident are unlikely to happen, as long as
appropriate radiation protection measures are taken in a timely manner;
however, all conceivable situations should be considered for the
response to a future nuclear disaster.
The evaluated MDA values (Figs. 9 and 10) are useful to estimate the
period of time during which direct thyroid measurements targeting 131I
are feasible, given its relatively short physical half-life and biological
half-life in particular for young children (ICRP 1989). Our new thyroid
monitor achieved an MDA value of ~30 Bq for a counting time of 180 s
under normal background conditions, which was comparable to that by
6


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

Fig. 10. MDA for the 131I thyroid contents as a function of the ambient dose
rate in the case of a measurement time of 180 s (Panel A for 4 × 1 detector
array, Panel B for 4 × 2 detector array).

a NaI(Tl) spectrometer with a 1-inch-dia., 1-inch thick crystal in similar
experiments (Yajima et al., 2020). It is notable that the total volume that
is sensitive to radiation is much smaller in the new thyroid monitor (4 or
8 cm3) compared to this NaI(Tl) spectrometer (12.9 cm3). On the other
hand, the MDA was ~40 Bq for a stationary-type thyroid monitor
equipped with a high-purity germanium (HPGe) semiconductor detector
mounted in a 50-mm-thick annular-shaped lead shield at our institute
(Kunishima et al., 2019; NIRS, 2016), although this monitor was difficult
to be applied to children.
Fig. 12 illustrates the TEDs corresponding to the MDA value for
different age groups as a function of the time after intake: Panel A for the
normal background and Panel B for the elevated background. The MDA
values are described in Section 4.3. Here we assumed that the physi­
ochemical form was elemental iodine and used the datasets of the agespecific thyroid 131I retention rates (Bq per Bq Intake) taken from the
MONDAL code (Ishigure et al., 2004) to prepare the figure. The MON­
DAL code has the database of retention/excretion rates for 42 radionu­
clides included in ICRP Publication 54 (International Commission on
Radiological Protection, 1989) and 78 (International Commission on
Radiological Protection, 1997) and was validated by comparisons with
the data on these references. Based on this result, it is expected that
thyroid exposure corresponding to 10 mSv in TED for the 1-yr-old age
group can be detected until about 25 days post intake under a normal
background level of 0.05 μSv h− 1 (or until 10 days post intake under an
elevated background condition at 2.5 μSv h− 1). The same estimation can

be applied to the other age groups using the figure, with the result that
the older an age group is, the longer the period is. These periods could be
regarded as an index for the time limitation of direct thyroid measure­
ments with the new thyroid monitor for each age group, although the

Fig. 11. Scenes of mockup measurements with the new thyroid monitor.

periods vary with the dose level of concern. The direct thyroid mea­
surements should be initiated in a timely manner as noted in the
Introduction; however, some delay would be unavoidable in scenarios in
which a significant release of radionuclides lasts >1 week and residents
living in the affected areas are then ordered to shelter indoors by au­
thorities during the release. The period available for the measurements
would thus be expected to be shortened (especially for young children),
and the prioritization of subjects should be considered using the above
estimations.
It is desirable that direct thyroid measurements are performed at
places with as low as possible ambient dose rates. The 2013 IAEA
guideline recommends that the ambient dose rate at the measurement
location be less than 0.2 μSv h− 1. However, this was difficult to imple­
ment in the screening campaigns conducted to identify the levels of
internal thyroid exposure to children at the end of March 2011, about
two weeks after the accident (Kim et al., 2012). Considerable elevations
of background radiation level can occur across vast territories after a
large-scale nuclear accident; indeed, this is considered one of the major
obstacles for the early initiation of direct thyroid measurements.
Finally, we would like to address the advantages of our new monitor
over existing devices. First, the unique probe shape allows stabilization
against the curved surface of the neck during measurements of young
children. Second, the multiple detector system allows identification of

irregular deposition of iodine in the thyroid (e.g., a difference between
the left and right lobes) if necessary. Third, the variation of the FEP
7


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

future.
6. Summary
We developed a new hand-held type thyroid monitor to measure
radioiodine in the human thyroid in the case of a major nuclear accident.
The main target subjects of this monitor are young children including
newborns. Multiple small GAGG detectors with 1-cm cube-shaped
crystals in each were arranged along the neck surface, and two detec­
tor arrangements (4 × 1 and 4 × 2 detector arrays) were finally selected
based on mock-up experiments. The narrow thickness of the 4 × 1 de­
tector array probe (24 mm) allowed for contact placement on the
anterior neck of the 1-yr-old mathematical phantom. The full energy
peak efficiency (FEP) of the monitor for 131I (at 365 keV) in the thyroid
was obtained by experiments using IRSN phantoms and simulations
using mathematical phantoms. Based on the results, minimum detect­
able activity (MDA) was evaluated for various ages. The MDA values for
subjects aged ≤5-yr-old were estimated as ~30 Bq at a normal back­
ground level for a counting time of 180 s. This value would be adequate
to detect the 131I thyroid content in 1-yr-old children corresponding to a
thyroid-equivalent dose of 10 mSv until about 25 days post intake. The
possible variation of the FEP values due to the displacement of the probe
was estimated to be about 20%. Our monitor would be useful for direct

thyroid measurements of young children following a major nuclear
accident.
Funding
This work was financially supported by the Nuclear Regulation Au­
thority of Japan under the Radiation Safety Research Promotion Fund
(JPJ007057).
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Fig. 12. Thyroid-equivalent doses (TEDs) corresponding to the MDA for
different age groups as a function of the time after intake via inhalation of 131I
in the form of elemental iodine. (Panel A for the normal background: 0.05 μSv
h− 1, Panel B for the elevated background: 2.5 μSv h− 1). The MDA values for
each condition are described in Section 4.3.

Acknowledgements
We would like to express our gratitude to Dr. Susumu Yokoya of
Fukushima Medical University and Dr. Nao Nishimura of National
Center for Child Health and Development for their valuable comments
on direct thyroid measurements of young children. We also appreciate
Drs. Tiffany Beaumont and David Broggio of IRSN for their support
regarding the development and delivery of the phantom sets and Dr.
Alexander Ulanovsky for providing MCNP’s input decks for his devel­
oped mathematical phantoms.

efficiency is expected to be small for children aged ≤5 yrs (Fig. 5). The
experimental FEP efficiency for the 5-yr-old IRSN phantom can be
applied to the FEP efficiency for younger children; the calculated peak
efficiencies for the 3-mo-old age group and the 1-yr-old age group were

about. 20% higher (Fig. 5), resulting in reasonable overestimations of
the 131I thyroid content.
We currently consider that the best use of our monitor is the addi­
tional or detailed measurements in combination with screening using
NaI(Tl) survey meters (e.g., TCS-171/172) at the venue designated on
the regional evacuation plan for each nuclear power plant site. Potential
subjects for the monitor are supposed to be persons (in particular, young
children) whose internal thyroid doses are found to be higher than a
certain dose level (e.g., 50 mSv in TED). Comparative measurements of
the same subjects by the two methods would surely be useful to validate
the results of the screening for dealing with large populations. Nonspectrometric devices would offer accurate determinations of the 131I
thyroid content based on appropriate calibration (Isaksson et al., 2019),
although one should evaluate the interference of other radionuclides in
measurements. In this regard, a dose rate meter recently developed for
thyroid monitoring is also of interest to improve the early response in a
nuclear disaster (Meisenberg and Gerstmann, 2017). We will describe
further considerations on the use of our new thyroid monitor in the

References
Beaumont, T., Ideias, P.C., Rimlinger, M., Broggio, D., Frank, D., 2017. Development and
test of sets of 3D printed age-specific phantoms for 131I measurements. Phys. Med.
Biol. 72, 4673–4693. 2017.
Beaumont, T., Rimliger, M., Broggio, D., Ideias, P.C., Frank, D., 2018. A systemic
experimental study of parameters influencing 131-iodine in vivo spectrometric
measurements using age-specific thyroid phantoms. J. Radiol. Prot. 38, 651–655.
/>Brenner, A.V., Tronko, M.D., Hatch, M., Bogdanova, T.I., Oliynik, V.A., Lubin, J.K.,
Zablotska, L.B., Tereschenko, V.P., McConnell, R.J., Zamotaeva, G.A., O’Kane, P.,
Bouville, A.C., Chaykovskaya, L.V., Greenebaum, E., Paster, I.P., Shpak, V.M.,
Ron, E., 2011. I-131 dose response for incident thyroid cancers in Ukraine related to
the Chernobyl accident. Environ. Health Perspect. 119 (7), 933–939. https://doi.

org/10.1289/ehp.1002674.
Broggio, D., Baud´e, D., Belchior, A., Berkovskyy, V., Bonchuck, Y., Dewogh
elaăere, J.,
Etherington, G., Fojtớk, P., Franck, D., Gomez-Ros, J.M., Gregoratto, D.,
Helebrant, J., H´
eriard Dubreuil, G., Hůlka, J., Isaksson, M., Kocsonya, A., Lebacq, A.L., Likhtarev, I., Lombardo, P., Lopez, M.A., Mal´
atov´
a, I., Marsh, J.W., Mitu, I.,
Monteiro Gil, O., Moraleda, M., Navarro, J.F., O´sko, J., P´
antya, A., P´
azm´
andi, T.,
Perez, B., Pospisil, V., Ratia, G., Saizu, M.-A., Sz´
ant´
o, P., Teles, P., Tymi´
nska, K.,

8


K. Yajima et al.

Radiation Measurements 150 (2022) 106683

Vanhavere, F., Vaz, P., Vrba, T., Vu, I., Youngman, M., Zagyvai, P., 2019. Child and
adult thyroid monitoring after a reactor accident (CAThyMARA): technical
recommendations and remaining gaps. Radiat. Meas. 128, 106069. />10.1016/j.radmeas.2019.02.008.
Cardis, E., Hatch, M., 2011. The Chernobyl accident − an epidemiological perspective.
Clin. Oncol. 23 (4), 251–260. />Currie, L.A., 1968. Limits for qualitative detection and quantitative determination.
Application to radiochemistry. Anal. Chem. 40 (3), 586–593. />10.1021/ac60259a007.

Gilmore, G., 2008. Practical Gamma-Ray Spectrometry, second ed. Wiley, New York,
ISBN 9780470861967.
Hosoda, M., Iwaoka, K., Tokonami, S., Tamakuma, Y., Shiroma, Y., Fukuhara, T.,
Mayo, Y., Taniguchi, J., Akata, N., Osanai, M., Tsujiguchi, T., Yamaguchi, M.,
Kashiwakura, I., 2019. Comparative study of performance using different gamma-ray
spectrometer for thyroid monitoring under emergency situations. Health Phys. 116
(1), 81–87. />Hosokawa, Y., Hosoda, M., Nakata, A., Kon, M., Urushizaka, M., Yoshida, M.A., 2013.
Thyroid screening survey on children after the Fukushima Daiichi nuclear power
plane accident. Rad. Emerg. Med. 2 (1), 82–86. .
jp/wp-content/files_mf/1465542161vol2_ rem_13_yoichirohosokawa.pdf.
International Atomic Energy Agency, 1988. The Radiological Accident in Goiˆ
ania. https
://www-pub.iaea.org/MTCD/Publications/PDF/Pub815_web.pdf.
International Atomic Energy Agency, 2013. Actions to Protect the Public in an
Emergency Due to Severe Conditions at a Light Water Reactor. EPR-NPP PUBLIC
PROTECTIVE ACTIONS, 2013. a.
org/MTCD/Publications/PDF/EPR-NPP_PPA_ web.pdf.
International Commission on Radiological Protection, 1989. Individual monitoring for
intakes of radionuclides by workers. ICRP Publication 54. Ann. ICRP 19 (1− 3).
International Commission on Radiological Protection, 1995. Age-dependent doses to the
members of the public from intake of radionuclides: part 4 inhalation dose
coefficients. ICRP Publication 71. Ann. ICRP 25 (3− 4).
International Commission on Radiological Protection, 1997. Individual monitoring for
internal exposure of workers. ICRP Publication 78. Ann. ICRP 27 (3− 4).
International Commission on Radiological Protection, 1998. ICRP Database of Dose
Coefficients: Workers and Members of the Public version 3.
Isaksson, M., Broggio, D., Fojtík, P., Lebacq, A.L., Navarro Amaro, J.F., O´sko, J., P´erez
L´
opez, B., Vu, I., Battisti, P., Bă
orjesson, J., Carlsson, M., Castellani, C.M.,

Gồrdestig, M., Hill, P., Krajewska, G., Lỹnendonk, G., Meisenberg, O., Stenstră
om, M.,
El Mantani Ordoulidis, S., 2019. Assessing 131I thyroid by non-spectrometric
instruments ‒ A European intercomparison exercise. Radiat. Meas. 128, 106115.
/>Ishigure, N., Matsumoto, M., Nakano, T., Enomoto, H., 2004. Development of software
for internal dose calculation from bioassay measurements. Radiat. Protect. Dosim.
109 (3), 235–242. />Japan Atomic Energy Research Institute, 1993. Health Physics in JAERI. JAERI-M-93172. />Khrutchinsky, A., Drozdovitch, V., Kutsen, S., Minenko, V., Khrouch, V., Luckyanov, N.,
Voillequ´
e, P., Bouville, A., 2012. Mathematical modeling of a survey-meter used to
measure radioactivity in human thyroids: Monte Carlo caluclations of the device
response and uncertainties. Appl. Radiat. Isot. 70, 743–751. />10.1016/j.apradiso.2011.12.032.
Kim, E., Kurihara, O., 2020. Thyroid doses in children from radioiodine following the
accident at the Fukushima Daiichi nuclear power plant. J. Radiat. Prot. Res. 45 (1),
2–10. />Kim, E., Kurihara, O., Suzuki, T., Matsumoto, M., Fukutsu, K., Yamada, Y., Sugiura, N.,
Akashi, M., 2012. Screening survey on thyroid exposure for children after the
Fukushima Daiichi nuclear power station accident. In: Proceedings of the First NIRS
Symposium on the Reconstruction of Early Internal Dose in the TEPCO Fukushima
Dai-Ichi Nuclear Power Station Accident. National Institute of Radiological Sciences,
Chiba, Japan, pp. 59–66. NIRS-M-252. />w_main&active_action=repository_view_ main_item_detail&item_id=73765&ite
m_no=1&page_id=13&block_id=21.

Kim, E., Kurihara, O., Tani, K., Ohmachi, Y., Fukutsu, K., Sakai, K., Akashi, M., 2016.
Intake ratio of 131I to 137Cs derived from thyroid and whole-body doses to Fukushima
residents. Radiat. Protect. Dosim. 168 (3), 408–418. />hp.0000000000001345.
Kim, E., Yajima, K., Hashimoto, S., Tani, K., Igarashi, Y., Iimoto, T., Ishigure, N.,
Tatsuzaki, H., Akashi, M., Kurihara, O., 2020. Reassessment of internal thyroid doses
to 1,080 children examined in a screening survey after the 2011 Fukushima nuclear
disaster. Health Phys. 118 (1), 36–52. />hp.0000000000001125.
Kramer, G.H., Crowley, P., 2000. The assessment of the effect of thyroid size and shape
on the activity estimate using Monte Carlo simulation. Health Phys. 78 (6), 727–738.

/>Kunishima, N., Tani, K., Kurihara, O., Kim, E., Nakano, T., Kishimoto, R., Tsuchiya, H.,
Omatsu, T., Tatsuzaki, H., Tominaga, T., Watanabe, S., Ishigure, N., Akashi, M.,
2019. Numerical simulation based on individual voxel phantoms for a sophisticated
evaluation of internal doses mainly from 131I in highly exposed workers involved in
the TEPCO Fukushima Daiichi NPP accident. Health Phys. 116 (5), 647–656. https://
doi.org/10.1097/hp.0000000000000995.
Kurihara, O., Kanai, K., Nakagawa, T., Takada, C., Momose, T., Furuta, S., 2012. Direct
measurements of employees involved in the Fukushima Daiichi nuclear power
station accident for internal dose estimates: JAEA’s experiences. In: Proceedings of
the First NIRS Symposium on the Reconstruction of Early Internal Dose in the TEPCO
Fukushima Daiichi Nuclear Power Station Accident. National Institute of
Radiological Sciences, Chiba, Japan. NIRS-M-252 :13–25. />ion=pages_view_main&active_action=repository_view_main _item_detail&item_id
=73765&item_no=1&page_id=13&block_id=21.
Likhtarev, I.A., Grulko, G.M., Sobolev, B.G., Kairo, I.A., Pră
ohl, G., Roth, P., Henrichs, K.,
1995. Evaluation of the 131I thyroid-monitoring measurements performed in
Ukraine during May and June of 1986. Health Phys. 69 (1), 6–15. />10.1097/00004032-199507000-00002.
Meisenberg, O., Gerstmann, U.C., 2017. Thyroid monitoring of adults and children after
reactor accident with a new dose rate measurement device. Radiat. Meas. 125,
150–153. />National Institute of Radiological Sciences, 2016. Activity Records of NIRS Employees in
Their Responses to the Fukushima Daiichi Nuclear Power Plant Accident. NIRS-M286. March 2016. (in Japanese). />ry_uri&item_id=73799&file_id=10& file_no=1.
Nishino, S., Tanimura, Y., Yoshitomi, H., Takahashi, M., 2020. Porotype test of a portable
thyroid dose monitoring using gamma-ray spectrometers. Radiat. Meas. 134,
106292. />Ohba, T., Ishikawa, T., Nagai, H., Tokonami, S., Hasegawa, A., Suzuki, G., 2020.
Reconstruction of residents’ thyroid equivalent doses from internal radionuclides
after the Fukushima Daiichi nuclear power station accident. Sci. Rep. 10, 3639.
/>Ulanovsky, A.V., Eckerman, K.F., 1998. Modification to the ORNL phantom series in
simulation of the responses to thyroid detectors. Radiat. Protect. Dosim. 79 (1− 4),
429–431. />Ulanovsky, A.V., Minenko, V.F., Korneev, S.V., 1997. Influence of measurement
geometry on the estimate of 131I activity in the thyroid: Monte Carlo simulation of a

detector and a phantom. Heath Phys 72 (1), 34–41. />00004032-199701000-00004.
Werner, C.J. (Ed.), 2017. MCNP User’s Manual. LA-UR-17-29981.
/pdf_files/la-ur-17-29981.pdf.
Yajima, K., Kim, E., Tani, K., Kurihara, O., 2020. A new thyroid monitor using multiple
high resolution Gd3(Al, Ga)5O12(Ce) detectors for direct thyroid measurements of
small children following a nuclear accident. Radiat. Meas. 133, 106272. https://doi.
org/10.1016/j.radmeas.2020.106272.
Youngman, M.J., 2013. Practical Guidance on Thyroid Monitoring for Radioiodine Using
Hand-Held Instruments. Health Protection Agency. HPA-CRCE-044. https://assets.
publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/f
ile/337139/HPA-CRCE-044_for_website.pdf.

9